Bridging Disciplinary Boundaries (January 11 - 14, 2007)


Pacific A (Hyatt Regency San Francisco)

Safe Spaces: Predictors of Children's Sense of Belonging in Afterschool Programs

Keith Warren, PhD, Ohio State University and Dawn Anderson-Butcher, PhD, Ohio State University.

Purpose: It is becoming increasingly clear that afterschool programs are not simply attempts to improve children's educational achievement; they also provide needed socialization and safe spaces for children (Halpern, 2006), and they are unlikely to do that unless children experience a sense of belonging (Anderson-Butcher & Conroy, 2002). This study analyzes the relationship between children's ratings of the quality of staff supervision and the safety of their environment and their sense of belonging at their afterschool program.

Methodology: 626 elementary and middle school students in 24 elementary and middle school afterschool programs were given program satisfaction surveys. The dependent variable, sense of belonging to the program, was measured using a four-item scale (Cronbach's alpha = .792). Children's ratings of the quality of staff supervision were measured using a six item scale (Cronbach's alpha = .845). Children's sense of personal safety was measured by a single item assessing their sense of being safe in the program. This variable was also subtracted from the children's assessment of the safety of their neighborhood, on the grounds that safety judgments could be comparative (Suls & Wheeler, 2000). Change in schools in the previous year, locus of control and academic level were also entered as control variables. A hierarchical linear model with program as the grouping variable and quality of staff supervision and academic level as group level variables was used in the analysis.

Results: As predicted, children's ratings of the quality of staff supervision were positively correlated with sense of belonging (t = 16.19, p < .001), as was sense of safety (t = 3.95, p < .001) and locus of control (t = 2.11, p = .03). Neither academic status nor change of school in the previous year predicted sense of belonging. Surprisingly, the difference between program safety and neighborhood safety was negatively correlated with belonging (t = -3.78, p < .001). A QQ plot indicated that residuals were acceptably normally distributed, and tests of independent variables as predictors of residuals did not indicate the presence of bias.

Implications: The positive relationship between individual children's assessment of program staff and their sense of belonging suggests that personnel may be as important as design in insuring program success. Children's sense of personal safety is also important, and it is likely that staff could do more to make afterschool programs safe spaces for students.